Compressed base morale of the units in EBII - why it is so?

[Jurand of Cracow]

Hi Guys,
As I compare data for the units in the EBII with those of the other mods, one thing looks very strange to me.

Namely, the EBII has a very compressed base morale of the units. I mean: there're very little differences accross the roster. Most of the units have between 2 and 6 morale, while the elites something like 7 or 8 (I've also seen 12 once - for Arioi, but Somatophylaces Strategou 7, Hyperaspistai 8 (I've checked it last time in 3.5). In comparison: in other mods the range goes up to 18, and the average units have 10 and the elites 16.
To my mind the effects are:
1. (result of the low morale, not the compression) units may flee earlier (ie. after having taken lower casualities), but then they regroup and come back to the battle (yes, they do, sometimes they regroup even under pressure of a chasing unit);
2. (result of compression) morale bonuses from the general are much more important (Command and Confidence raise the base morale a lot that means the threshold when the units start fleeing is much higher). As a result, a great general is absolutely priceless: my normal generals add between 4 and 7 to morale from various sources, while my experienced FL: +6 Command, +6 CommandWhenAttacking, +6 Morale, +5 Confidence - this sums to a whopping +23);
3. (result of compression, marginal effect) bonuses from the neighbouring units slightly more important (they add +1 to the friendly ones morale (eg. "command"), or take away 1 from the enemy (eg. "frighten cavalry");
--> as a result, in practice there's no serious difference if a unit has a good morale or not - everything hinges upon the general. This means there's little difference between crappy and good units. For the battle gameplay, this limits the tactical possibilities of playing - you don't make choices on the basis of morale as all the units are similar.
So my question is: why the EBII doesnt' use the whole spectrum of possible morale (1-18)?
Or maybe I'm wrong in the analysis?

Spoiler Alert, click show to read: 

[Neyak]

This probably only partially answers your question, but note that whether or not a unit will break etc. depends also on their discipline, not just the pure morale value. Elites, for example, have a morale value that's only a couple of points higher than skirmishers, but they are way less likely to break (lose fewer morale points or whatever the implementation details of the system are) under a heavy cavalry charge for example because of their discipline, so the morale points that they do have mean more.

And at least from my experience elites etc. stay significantly longer in a losing fight, for example, while some light archers get charged by light cavalry, lose 10% of their men and make a run for it... More often than not the human player is much better than the AI at avoiding, say, cavalry charges from the rear, so you might get the better data by looking at the order in which the AI troops rout. Often enough the battle was won except those enemy epilektoi hoplitai kept fighting and fighting and fighting for another ten minutes after everything else was said and done.

[nvm]

You should include Halithoz (the bodyguard) in your comparsion. These guys...I don't know what their morale is, but I've literally never witnessed them break, even when completely surrounded and fighting as the last unit on the battlefield.
Also regardless of the stats I found the morale divide to be pretty significant between levies and elite units, though I can agree generals stats likewise seem to matter a lot. There's a huge difference between fighting battles with an experienced generals and completely new ones. (That is, I end up making manuevres which would be fine with a decent generals, but result in troops breaking early very quickly using a fresh one) Once a general gets some experience it tends to mostly even out, up until to dealing with truly great generals like Pyrrhos.

[Jurand of Cracow]

This probably only partially answers your question, but note that whether or not a unit will break etc. depends also on their discipline, not just the pure morale value. Elites, for example, have a morale value that's only a couple of points higher than skirmishers, but they are way less likely to break (lose fewer morale points or whatever the implementation details of the system are) under a heavy cavalry charge for example because of their discipline, so the morale points that they do have mean more.

And at least from my experience elites etc. stay significantly longer in a losing fight, for example, while some light archers get charged by light cavalry, lose 10% of their men and make a run for it... More often than not the human player is much better than the AI at avoiding, say, cavalry charges from the rear, so you might get the better data by looking at the order in which the AI troops rout. Often enough the battle was won except those enemy epilektoi hoplitai kept fighting and fighting and fighting for another ten minutes after everything else was said and done.

I'm uncovinced: discipline is important only in specific moments for it is the resilence of the unit to a morale shock (general killed, sudden loss of men - what else?). Given that sudden loss of men happens more rarely in the EBII than in the other mods (because of the low killing ratios), it doesn't matter much.
The example with the archers is a cheap shot - this should and will always happen. I have melee men in mind, starting with tribal infantry up.

Also regardless of the stats I found the morale divide to be pretty significant between levies and elite units, though I can agree generals stats likewise seem to matter a lot. There's a huge difference between fighting battles with an experienced generals and completely new ones. (That is, I end up making manuevres which would be fine with a decent generals, but result in troops breaking early very quickly using a fresh one) Once a general gets some experience it tends to mostly even out, up until to dealing with truly great generals like Pyrrhos.

Levies vs. elits: 3 vs. 8. While in the DIK or SSHIP is 5 vs. 15. So the differences are 5 and 10 - and in the case of morale the nominal differences is important, not percentages.

[Neyak]

Had to check ... the Halithoz have morale 12, and are obviously "disciplined". I think that's the highest there is. The lowest is 2 for infantry (many light skirmishers and e.g. pantodapoi, who will run if someone on a horse just looks at them funny). The lowest is 1 for cavalry (low-end skirmisher cavalry), but in practice the latter interact differently with morale anyway, so that's not really a good comparison.

[Neyak]

I'm uncovinced: discipline is important only in specific moments for it is the resilence of the unit to a morale shock (general killed, sudden loss of men - what else?). Given that sudden loss of men happens more rarely in the EBII than in the other mods (because of the low killing ratios), it doesn't matter much.
The example with the archers is a cheap shot - this should and will always happen. I have melee men in mind, strating with tribal infantry.

Fair enough, let's just focus on line infantry of various quality. One thing I think that is relevant here where discipline comes into play is being flanked. And you can get routs on lighter infantry by flanking them with other infantry. However, if it's really just line-vs-line you'd expect this to last for quite a while even if unequal opponents meet. You might expect a gradual retreat by the weaker party. I'd be hesitant to engage someone with a sword and a cape and a bunch of armour if I had but a tunic and a spear (I'd probably hang out cowardly in the vicinity and try not to get killed until the whole nightmare is over), but an outright rout would need to be sparked by something, e.g. the guys next to me routing first, or having enemies appear from an unexpected direction.

Part of the issue, I think, is that there is no "organized retreat" mechanism, by which a unit might slowly disengage / move backwards but at least maintain some degree of discipline.

[QuintusSertorius]

It's primarily reason 1, which changes the dynamic of battles. Lower-morale infantry might rout quickly, but will do so with a large proportion of it's numbers intact. Meaning if it gets a chance to rally, it will return to the fight sufficiently whole to still pose a threat. At a higher morale point (beyond about 12) modifiers are all but irrelevant. When a unit finally breaks at a higher morale point, it's so shattered there's no chance of rally, and if it should happen to do so it poses no threat at all.

I don't agree there's no difference between units, there's a pretty significant difference between a levy with 1 or 2, and a solid medium with 5 or 6. Also note cavalry have deliberately lower morale than infantry to ensure they will withdraw when presented with shocks. Light cavalry can break, rally and return repeatedly before they're done.

[Lusitanio]

I'm with Quintus here. You can feel a clear difference with moral between units, especially between levies and elites. Besides that, battles in EBII are very well balanced, there is no need for further change regarding moral.

[Jurand of Cracow]

Thanks, Quintus, for the answer!

I.
I do understand the rationale now: the morale values for good units cannot be higher because if they were, these units would never return to the battlefield to give a good fight.
I do appreciate low morale values in the EBII, and also the very low ones for the cavalry units - these are indeed realistic, imo.
I'm not convinced that there's a significant difference between a levy with 1 or 2, and a solid medium with 5 or 6. It's too low to make any change, imo.

The question is if the trade-off is worth it. In the first version, we could have big differences between the units (what I appreciate from the point of view of tactical play of the battles - the more tactical choices, the better). The drawback of such situation is that the good units wouldn't return to the battefield once they're routed.
In the second version, we have compressed base morale (less differences between units what I find it inferior) but they come back and fight even they're routed (yes, I can confirm it, I've many AI units coming back and fighting well).
The EBII team choice was for the latter. I understand now why.

II.
Could you tell us what are your thoughts concerning the generals' impact on the morale?
I find those high values (and I do record data for my generals - they always have at least +3 to morale, and usually around +7 - from different sources) negating the risks for the player to break on morale. I think it adds much to the compression of morale.
I think the EBII is too lavish on the various morale sources, ie it doesn't grant too many from one source, but they add up from many sources:
- Command - I usually get 1-2 stars for normal generals, while the leading one (FL) gets up to 6 after 10 battles or so;
- Command while Attacking - Pritanoi get it often, and these matters as it's the player who usually attacks (Command while Defending is more important for the AI),
- Confidence (what used to be Chivalry in vanilla, but without Population Growth effect) - usually between 2 and 4;
- TroopMorale - again a few points for a warring general.
My opinion would be: make all TroopMorale translate into Confidence (for now, there's a mess, as discussed here).

Spoiler Alert, click show to read: 

III.
BTW, do you know any info similar to the data on morale impacts as stated in this thread (or maybe it's right also for the M2TW)?

IV.
On the powerful charge - I'm not sure if we're getting right what's the effects are. I've asked once PointBlank, who - afaik - created the system for the Stainless Steel, and he provided the following information:

The attribute power_charge doesn’t do anything. Otherwise Powerful Charge appears on the unit card if the unit’s charge or attack + charge, never checked which (maybe it factors in mount mass too) exceeds some threshold.
So can_formed_charge doesn’t really link to Powerful Charge

power_charge is non-functional, it has zero effect.
Heavy Cavalry is just a designation for use by the AI when constructing formations, and as a cue to the developer when deciding on which voice_type to use.
The charge effect for all cavalry is derived from a combination of their attack, charge value, mount mass, whether they are running, if they are in a 'proper' charge, their x-radius, the global melee_hit_rate as defined in Battle_config.xml, and to some degree their animation, which is different for couched and non-couched lances for example.

I'm not sure if Point Blank is right. However, I recall that adding power_charge attribute in EDU to a cavalry unit didn't result in the unit gaining it (at least it didn't appear in the in-game card description).
Furthermore, what's the role of the "can_formed_charge"?
In the EDU description the adnotation is "power_charge = unkown". The "can_formed_charge" is not mentioned. However, in the TWC guide to EDU you've got descriptions "power_charge – Seems to improve the charging ability of a unit" and "can_formed_charge – The unit can form up and charge".

The SSHIP EDU - made by MWY, another guy I consider to be expert in the matter - only "can_formed_charge" is being used (no "power_charge") and it works well, in my experience.
The DIK EDU - the same, only "can_formed_charge" - and only for the cavalry.
The TATW-DaC EDU - uses both "can_formed_charge" and "power_charge" - but never simultaneously.
The Dawn of Conquest - only "can_formed_charge", but only for bodyguards - regular knights don't have any...

In the EBII EDU only "power_charge" is used, no "can_formed_charge".
But then things get complicated:
[unit --- sum of attack and charge --- EDU "power_charge" --- in-game description Powerful Charge]
Arioi (inf) --- 26 --- yes -- yes
Uxselouiroi Desnouon (inf) --- 22 --- yes --- no
Nedes Nesamoi (inf) --- 22 --- yes --- no
Batoroi (inf) --- 20 --- yes --- no
Magones (cav) --- 23 --- no --- yes
Eqoreda (cav) --- 19 --- no --- yes
Equites Romani (cav) --- 19 --- no --- yes

All in all, I'm all at sea.

[Rad]

I had my doubts when the cavalry morale was drastically lowered. However, a few campaigns convinced me that it was a good decision.
If caught in a tight spot, cavalry runs instead of taking a beating. Most of the time, it reforms and continues the fight - seeing enemy cavalry getting chased away from the battlefield has become a thing I truly enjoy. Cavalry gets to reform because it's fast and has time to recover morale. It will rarely get mowed down by pursuers.

I'm not so sure that low morale works well for infantry. Not for the AI, at least.

Sure, the player can counter the low morale of the infantry by having a good general and not exposing vulnerable units, but the AI rarely does that. Despite the fact that I end most battles after all the enemy units are routed, I have tons and tons of prisoners after each one, especially against factions that have plenty of low morale foot skirmishers, like Numidia does. Hell, playing as Carthage, I even routed a full Camilian Roman army by using an echelon formation. I placed my best troops on the right. They hit first, they hit fast. The Roman left was smashed to bits. That started a chain reaction. The rest of the Romans rolled up like a carpet in 2 minutes, my left and center barely made contact with them before they ran. The result? I killed 200-300 Romans, the rest were captured. I lost fewer than a hundred soldiers. I couldn't believe it. Like I said, it was a full Camilian army, packed with Roman heavy infantry and with cavalry on the flanks. One of the rare instances I was satisfied with the AI Roman army composition. My army wasn't bad, but it was not what I usually take to war against Rome. They were pacifying the locals in southern Gaul and got caught up. I actually expected to lose that battle or at least suffer extremely heavy casualties - that's why I opted for the echelon formation. I figured that I could at least delay a chain rout in my army by having my weakest units engage last. It turned out to be my greatest victory in any med2 mod, ever. Was it fun? Heck yeah! Was the outcome historically reasonable? Yes. It had Leuktra spelled all over it. Does it fit in well in what is after all, a game? I am not so sure. The AI can't effectively deal with the issue of having low morale infantry.

In my experience, infantry charges are underwhelming in this mod. Sure, I don't expect them to break phalanxes like they were twigs, but I expect something. The preceding javelin volley has a far greater impact, which is why I charge after all my ammo is spent.




edit: Stannis

[nvm]

Perhaps morale for most infantry units could be slightly bumped, with skirmishers remaining the same. Most of what makes troops capable of holding a battle line comes from generals. Of course AI doesn't tend to have good generals often, in fact usually they won't have one at all which makes swatting their armies away effortless.

[Rad]

Yeah, it's captains almost half the time.

[QuintusSertorius]

There's nothing complicated about the output on the unit card, it's garbage. You're better off ignoring it altogether, none of it is meaningful. Only what's in the EDU matters.

[Jurand of Cracow]

There's nothing complicated about the output on the unit card, it's garbage. You're better off ignoring it altogether, none of it is meaningful. Only what's in the EDU matters.

I didn't mean the unit card info is the problem. The unit card is just a symptom of the illness, not the illness itself or it's consequences.

I can discern two consequences of the problems and their underlying causes:
1) the morale mechanism in the EBII doesn't work as intended. It is meant to be low in order to enable the rout-recover-return process. However, in practice (given the many sources of morale I've enumerated in my post) the player's units get very high morale (irrespectively if they're elite or not). The player is thus in a superior position and he cares much less about the rout of his units. This is because any reasonabe player will have his armies commanded by a good general who would prevent such a rout. Additionally, it makes the normal units similar to the elite units - it's the "compressing" effect of those high morale buffs from the general. It's EDCT what matters and what's coded there.
2) the charge mechanics doesn't work as it supposedly intended (for the infantry, at least). The number of kills in this phase is very low and - given the EBII low kill ratio - this phase is very short (as a part of the whole fight). One of the consequences is that the historical advantage of the barbarian units - the punch at the beginning of the fight - doesnt work. My point is that the reason is the errneously coded attribute in the EBII EDU: attribute "power_charge" that reportedly doesn't do anything. I fully agree that it's EDU that matters.

So my questions are:
1) Could you tell us what are your thoughts concerning the generals' impact on the morale? Is it intended that that the player - in practice - have armies that don't break in the weaker segments (they obviously may break if overwhelmed by the enemy but there's a little difference between good and weak units in the respect of morale)?
2) Does the charge mechanics work as intended? Is coding "power_charge" in the EDU not an error in coding? Shouldn't "can_formed_charge" be used instead? Or they both doesn't matter? Then maybe it's better to simply scrap it? Or we simply don't know (then why not to ask eg the TATW team). Or I'm wrong in my analysis - then where's the error in reasoning?

[QuintusSertorius]

Relying on the unit card for anything is a problem, because the output is nonsense. Discount it from your analysis altogether.

The general's impact, in proper battles (not auto-calc) is limited in range (the size of the radius depends on their Command). It affects the units closest to him, not all units everywhere on the field. The number of elites with "command" is more important in this respect than the general alone. The AI's inconsistency in using proper commanders is a bigger issue than the values you might be able to stack on a single FM if you're trying.

Charging is a completely different matter, and it doesn't work properly because the M2TW engine isn't designed with precursor weapons in mind. That the prec attribute is there at all is a vestigial feature from RTW they didn't bother removing. It still works better in 2.35 than it did in 2.3 when we'd removed prec altogether.

I'd need more evidence than just someone's say-so on whether or not power_charge works.

[Jurand of Cracow]

Relying on the unit card for anything is a problem, because the output is nonsense. Discount it from your analysis altogether.

The general's impact, in proper battles (not auto-calc) is limited in range (the size of the radius depends on their Command). It affects the units closest to him, not all units everywhere on the field. The number of elites with "command" is more important in this respect than the general alone. The AI's inconsistency in using proper commanders is a bigger issue than the values you might be able to stack on a single FM if you're trying.

Charging is a completely different matter, and it doesn't work properly because the M2TW engine isn't designed with precursor weapons in mind. That the prec attribute is there at all is a vestigial feature from RTW they didn't bother removing. It still works better in 2.35 than it did in 2.3 when we'd removed prec altogether.

I'd need more evidence than just someone's say-so on whether or not power_charge works.

Unit cards - I don't know whether it's just rubbish, but in the other mods (and I've played many) the units with Powerful charge were really good at charging. This was something I'd always pay attention to for it was a good indicator of the abilities of the unit. But as said - it's a symptom and should be treated as just a symptom, so discounting from the analysis is ok.

I recall to have very good infantry charges playing in various M2TW mods, especially with heavy axe units. The charges were not as deadly as with heavy cavalry, but still much decisive. What I understand is that the EBII team concluded that neither with the precursor weapons nor without the it doesn't work properly. Maybe it's the problem of the M2TW engine, but in my opinion is heavily related to the low kill ratio of the EBII (and thus a short time of application of the charge value).

Indeed, a part of the general's traits impact in the limited range. However, the TroopMorale is applied to the all units on the battlefield. Furthermore, the more Command has, the larger the range. I don't know the exact values, but with the high values (and look at the table with my generals: Command plus Command While Attacking) they cover a large chunk of the battlefield (I don't know how big, though). I assume that my king provides his huge morale boost for all the troops - and this is not unique, I'd expect other players also to have great kings giving so much morale.

Actually, Point Blank is not just "someone" but the guy who created the whole Real Combat system for the Stainless Steel. I would consider him to be one of just a handful guys ever who knew the ins-and-outs of the M2TW combat engine (though not the other features of the M2TW). Therefore I wouldn't dismiss his opinion so swiftly. Plust I've just mentioned that the other mods around don't use "power_charge" - with intriguing case of TATW using both power_charge and can_formed_charge.

[QuintusSertorius]

Unit cards - I don't know whether it's just rubbish, but in the other mods (and I've played many) the units with Powerful charge were really good at charging. This was something I'd always pay attention to for it was a good indicator of the abilities of the unit. But as said - it's a symptom and should be treated as just a symptom, so discounting from the analysis is ok.

No, unit cards report nonsense. Heavy cavalry units have "fast moving", yet light cavalry units don't? 0s reported as 1s, because you can't have a zero? The list goes on, they are not reliable and shouldn't be used in any analysis. Only direct comparison of the EDU.

I recall to have very good infantry charges playing in various M2TW mods, especially with heavy axe units. The charges were not as deadly as with heavy cavalry, but still much decisive. What I understand is that the EBII team concluded that neither with the precursor weapons nor without the it doesn't work properly. Maybe it's the problem of the M2TW engine, but in my opinion is heavily related to the low kill ratio of the EBII (and thus a short time of application of the charge value).

Did those units have primary missile weapons? I'm not talking about the prec attribute, but having javelins as a primary weapon, something the engine clearly isn't designed to do. However, which is something pretty ubiquitous in the period covered by the game.

Indeed, a part of the general's traits impact in the limited range. However, the TroopMorale is applied to the all units on the battlefield. Furthermore, the more Command has, the larger the range. I don't know the exact values, but with the high values (and look at the table with my generals: Command plus Command While Attacking) they cover a large chunk of the battlefield (I don't know how big, though). I assume that my king provides his huge morale boost for all the troops - and this is not unique, I'd expect other players also to have great kings giving so much morale.

TroopMorale is hardly used in the EDCT compared to Command (the latter appears more than 6 times more frequently), and most of the references are negative. So for the most part we are not talking about global bonuses for the player. Or indeed the AI, since AIGeneral doesn't have TroopMorale.

Command has a limited radius, the only advantage the player has is if they're better at getting their commander to the trouble spots. Or are targeting the enemy commander every time to get easy wins (one of the reasons I've given every Rebel FM HaleandHearty for extra hit points, and boosted the hit point bonus for AIGeneral in the patch).

Actually, Point Blank is not just "someone" but the guy who created the whole Real Combat system for the Stainless Steel. I would consider him to be one of just a handful guys ever who knew the ins-and-outs of the M2TW combat engine (though not the other features of the M2TW). Therefore I wouldn't dismiss his opinion so swiftly. Plust I've just mentioned that the other mods around don't use "power_charge" - with intriguing case of TATW using both power_charge and can_formed_charge.

Again, I'd like to see some evidence.

[Jurand of Cracow]

No, unit cards report nonsense. Heavy cavalry units have "fast moving", yet light cavalry units don't? 0s reported as 1s, because you can't have a zero? The list goes on, they are not reliable and shouldn't be used in any analysis. Only direct comparison of the EDU.

I agree that unit card data are not reliable for analysis. I recall that "fast" attribute depends on the mount type (and perhaps definition of the mount etc.). I'm just saying that some symptoms may be infered from the description in the unit cards. I detect differences in the charges between units with or without words "Powerful charge" and the in-battle performance of the units, but it may just be me. Anyway, given my interests in the details, I'm using the the Extended Info submod (and obviously even this approach is imperfect as animations or mounts are also important).

Did those units have primary missile weapons? I'm not talking about the prec attribute, but having javelins as a primary weapon, something the engine clearly isn't designed to do. However, which is something pretty ubiquitous in the period covered by the game.

No, they didn't. And I agree that the M2TW engine is not perfect for the EB timeframe, also in this respect (lack of sapping is another issue).

TroopMorale is hardly used in the EDCT compared to Command (the latter appears more than 6 times more frequently), and most of the references are negative. So for the most part we are not talking about global bonuses for the player. Or indeed the AI, since AIGeneral doesn't have TroopMorale.

Command has a limited radius, the only advantage the player has is if they're better at getting their commander to the trouble spots. Or are targeting the enemy commander every time to get easy wins (one of the reasons I've given every Rebel FM HaleandHearty for extra hit points, and boosted the hit point bonus for AIGeneral in the patch).

Ok, fair play. I understand the approach, this is a legitimate choice (even if I think that all TroopMorale should translate into Confidence for clarity, and there should be less bonuses altogether to limit that "compression").

Again, I'd like to see some evidence.

If I come across, I'll put the link here. Me and Rad reported on non-existence of the charge effects (at least for the infantry) and I quoted the other mods (but this may be the herding effect).

Or maybe somebody from the other players could conduct tests?

[Ngugi]

Interesting; not a matter I've considered [merely adhered to King Kong's settings over at TATW] but migth be can_formed_charge keeps the pace/pack units tighter: http://www.twcenter.net/forums/showt..._formed_charge
Would require some comprehensive testing, it seems, on both attributes, since all I find on power_charge is presumptions?
Considering the EDU description of can_formed_charge I would not be surprised if its for cavalry only.
For the record can_formed_charge is not in Rome 1 or BI, so it is not a left over attribute (as sometimes occur). power_charge however is in BI.

[z3n]

Interesting; not a matter I've considered [merely adhered to King Kong's settings over at TATW] but migth be can_formed_charge keeps the pace/pack units tighter: http://www.twcenter.net/forums/showt..._formed_charge
Would require some comprehensive testing, it seems, on both attributes, since all I find on power_charge is presumptions?
Considering the EDU description of can_formed_charge I would not be surprised if its for cavalry only.
For the record can_formed_charge is not in Rome 1 or BI, so it is not a left over attribute (as sometimes occur). power_charge however is in BI.

I believe formation_hold_distance in the pathfinding file matters somewhat more than any of those (can_formed_charge / power_charge). Whenever I tweaked that I noticed significant effects.